The present invention relates generally to computer system errors, and more particularly to increasing the efficiency of platform dumps when a major error occurs that the computer system cannot process.
Typically, on computer servers, when a major error occurs which the system cannot resolve or progress from, sometimes called a hardware or software checkstop, a platform data dump collection, also known simply as a dump, happens, which assists in failure analysis. The data dump requires various services to be locked from use during the dump. Locking the various services is needed to resolve concurrent hardware access issues, and generally causes the system to be unusable while the platform dump data collection occurs. Additionally, the collection may take upwards of several minutes to complete.